L-Acoustics' Source Intelligence uses AI to separate vocals live at Céline Dion's Paris shows
Original title (Chinese)
AI 实时分离人声,席琳 · 迪翁演唱会用上音频黑科技
AISummary
L-Acoustics has released Source Intelligence, a machine-learning software platform that separates a singer's voice from other instruments, backing vocals, crowd noise and reverb in real time.
The software can cut non-vocal signals by up to 40 dB and works only with L-Acoustics speakers and its L-ISA Mk2 processor.
Source Intelligence is being used for Céline Dion's 26 residency shows at Paris La Défense Arena, which the source describes as a stadium with 45,000 capacity.
Source: IThome · AI · ithome.comPublished · added here